In this paper, I dissect the semantic assertion that keywords are polysemous by approaching keywords within a model of semantic change. From a linguistic perspective, the general use of the term polysemy in the discourse surrounding keywords is contentious; I argue that keywords tend to be more vague than polysemous.

I apply specific linguistic polysemy tests to corpus examples of keywords to show that the contested meanings of keywords are not generally constant to the word (polysemic) but contingent upon contextual enrichment (vague), a distinction that matters because fluctuating contextual enrichment is definitive of an ongoing, unsettled contestation of meaning. In order to explain why the vagueness in keywords does not tend to resolve into stable polsyemy, I refer to sociolinguistic notions of performative identity and accommodation theory to propose that social and cultural debate is a generalizable, pressurizing force that characterizes keywords and impacts semantic change.

This paper, then, fills a gap in Keywords research by bringing semantic mechanisms to the forefront, and then linking those mechanisms to the nature of social and cultural debate.

The question of whether it is possible to ‘tell the same story twice’ has been explored in work on conversational narratives, which has set out to understand the existence of some kind of ‘underlying semantic structure’ and ‘script (Polanyi, 1981). In this speech event, ‘local occasioning’ and ‘recipient design’ (Sacks, Schegloff and Jefferson, 1974) are factors that determine the form and function of the story as ongoing talk frames the narrative while other participants provide a ready made audience, all of which, form part of the storytelling process.

What happens, however, when a survivor of 7/7*, whose personal narrative was reported globally on the day of the event, is again interviewed two and a half years later for their experience of that morning? Is the “same story” retold? Specifically, how far does the latest story replicate the experience of the first and which of the prototypical features of a personal narrative – at the level of both the macrostructure and microstructure – remain constant? By comparing both interviews and using Labov and Waletzky’s (1967) narrative framework as the central model for analysis, it is possible to see whether events within the complicating action or features of evaluation remain the most memorable and are the core narrative categories. While findings show that both narratives are comparable in form, a closer investigation finds compelling differences as well as unexpected linguistic choices. Not only has the second narrative become informed by other, external narratives to become part of a broader, mediated narrative but various discourse strategies of ‘dissociation’ in both interviews have resulted in a retelling of a traumatic experience that appears to be closer to an eye witness report than a personal narrative. Moreover, this blurring of two distinct genres of storytelling provides a true insight of how the narrator positions himself inside this terrible experience.

*On July 7, 2005 there were a series of co-ordinated terrorist bomb attacks on the London transport system in the morning rush hour. Three bombs exploded on London underground trains (just outside Liverpool Street, Edgware Rd and Russell Square stations) and one on a double-decker bus in Tavistock Square, in central London. The bombings killed 52 commuters, the four suicide bombers and injured many hundreds of people.